Conservation of terrestrial invertebrates fundamentally depends on robust data. Pitfall trapping is a time- and cost-efficient method that provides valuable information. There is limited research on how pitfall trap colour influences catch, although colour is known to be important for other trap types, including pan traps. Here, we use data from 2,897 trapping days in grassland, parkland, farmland and woodland habitats to explore effects of pitfall trap colour (blue/orange/green) on catch abundance (number of invertebrate individuals caught), diversity (Shannon’s index), and community composition. We also assess how trap catch is influenced by trapping period (daytime vs. nighttime) and weather (precipitation and temperature). Catch abundance and diversity was higher in green traps than blue or orange traps overall and in grassland or parkland, whereas there was no effect of trap colour in farmland and woodland. For specific orders, catch abundance was significantly higher in green traps for Araneae, Coleoptera and Isopoda, but higher in orange traps for Hymenoptera in parkland and farmland, and higher in blue traps for Stylommatophora and Coleoptera in woodland. There was a significant time of day * colour interaction: green traps yielded higher catch abundance and diversity than blue or orange traps for day and night trapping, whereas orange traps yielded higher catch abundance and diversity than blue traps for night trapping only. As expected, catch abundance and diversity was higher in warmer temperatures, while precipitation had taxa-specific effects on catch abundance (Stylommatophora = positive; Isopoda = negative). There are opportunities to optimise protocols for specific habitats and taxa. Based on the results of this study, which has considered four sites (each covering a single habitat: grassland, farmland, parkland, woodland) across two years, we suggest that conservationists using pitfall traps for a census of invertebrate communities should use green traps if selecting a single colour, or a combination of colours. For more focused projects, we offer taxon- and habitat-specific recommendations that are achievable without additional resource demands.

Robust and reliable data underpin mammalian conservation, including priority setting, informing interventions, and measuring success. However, because many terrestrial mammals are nocturnal, crepuscular, or cryptic, obtaining data can be challenging and expensive. Problems are often magnified when focal species are rare or occur at low density and methods that involve direct intervention or create disturbance can have ethical implications. Use of non-invasive or minimally-invasive methods to survey mammals is not new. However, such approaches are continually developing and although there have been reviews for particular mammalian guild/location combinations, or for specific techniques, to date there has been no overarching review within the specific context of mammal conservation. In this review, we evaluate innovative techniques with broad applicability across mammalian taxa, habitats and contexts that have recently advanced substantially: technological approaches (camera trapping, bioacoustics, thermography, unmanned aerial vehicles); indirect methods (footprints, hair), novel approaches (detection dogs; eDNA metabarcoding), and participatory methods (technological citizen science, iEcology). We consider recent advancements, residual challenges, and take a horizon-scanning approach to highlight future development, empirical testing, and optimisation needs. Key findings include: advances in scalability of novel techniques are often due to development of open-source devices and programming interfaces; use of AI for species identification and quantification is already revolutionizing workflows; and rapid ongoing development of statistical models and analysis tools is broadening data processing potential and elevating utility of data collected using non/minimally-invasive methods beyond presence-only applications. We encourage practitioners embrace emerging tools in the conservation toolbox and to continue developing novel techniques.

Fenced reserves provide benefits for wildlife but are vulnerable to overgrazing. Accurate veld condition assessments are essential to inform stocking density, translocation/culling interventions and burn management decisions. Here, for the first time, we test whether atypical changes in the grass community at the edges of burn management blocks confound the accuracy of veld condition scores. Over a two-year period, 618 sequential parallel transects were conducted 0-240 m from edges of burn blocks and, for comparison, at the core (> 1 km from edge) of burn management blocks in a fenced reserve in South Africa. Veld condition was calculated using the Ecological Index Method. In areas burned recently (similar to 10 months previously), veld condition scores were unrepresentative within 60 m of the edge due to significant changes in grass community composition (i.e. unrepresentative condition scores were driven by atypical grass communities). In established areas (similar to 4 years since burning), no edge effect was found for veld condition scores, despite grass composition varying significantly (i.e. condition assessment was robust to community change). For simplicity and consistency, we recommend veld assessments are always conducted > 60 m from edges and that managers consider consequences of reduced veld condition at the edges of recently burned blocks, which could sum to substantial areas.

Ornithologists use many methods to advance understanding of birds and inform conservation strategies. These include field methods (e.g. censusing, ringing), laboratory analysis (e.g. molecular taxonomy, microbiology) and computational approaches (e.g. population modelling, habitat suitability mapping). Indirect metrics can also be important: for example, ptilochronology as a proxy for food availability and dominance hierarchies, stable isotopes as a proxy for diet and first egg dates as a proxy for phenological responses to climate change. Recently, the potential to use palynology – the study of pollen and spores – in ecology has become increasingly recognized, but there has been no specific appraisal of the utility of palynology in ornithology. In this scoping review, we synthesize research undertaken over 70 years across all seven continents covering the many diverse ways in which studying pollen associated with birds can advance ornithological understanding and inform conservation strategies. We critique the use of palynology both directly and as a proxy to study avian habitat use, migration ecology, diet and feeding strategies, and population dynamics, offering constructive suggestions for future research at the end of each thematic section. We also explore the potential for analysing pollen on birds (or within avian deposits such as guano) to answer wider questions, especially in relation to pollination and environmental reconstruction, thereby offering cross‐disciplinary collaborative research ideas. We conclude that the potential for uniting ornithology and palynology within the research and conservation contexts is as extensive as it is exciting and urge avian biologists to think broadly about new opportunities.

Passive acoustic monitoring (PAM) is an important survey method used to collect data on bat distribution and activity that are needed to underpin conservation and management. The introduction of open-source acoustic recorders at price points considerably lower than those for commercial detectors means PAM is becoming increasingly accessible to practitioners. However, uncertainty regarding recording quality, especially at higher frequencies, makes understanding the comparative performance of commercial and open-source devices imperative. Here, two types of commercial bat detectors: full spectrum (Anabat Swift) and zero-crossing (Anabat Express), and open-source AudioMoth acoustic recorders (configured to use 250 kHz (hereafter low) or (384 kHz hereafter high) sampling rates) are compared in each of four different habitats: riparian, woodland, wood pasture and arable. In each habitat, comparisons are made using detectors that were spatially co-located and recording on the same nights, such that they had identical opportunity to record the same data. Species accumulation curves were additionally created for each detector type to quantify the combined effects of using multiple detectors, multiple locations per site and multiple temporal replicates. When directly compared, full spectrum commercial detectors outperformed open-source devices (regardless of sampling rate) for many metrics, including for species richness (i.e. the number of species recorded) in all habitats tested. However, the low frequency open-source device performed similarly to the zero-crossing commercial device in quantifying overall bat activity and activity of individual taxa in most habitats. The low frequency open-source unit consistently outperformed the high frequency open-source unit. Commercial detectors accumulated species more quickly, and one detector at one location per site was typically sufficient to record the full species inventory. However, use of multiple open-source devices over a longer period recorded the same species inventory in three of the four habitats. Practical implication: Although commercial devices remain the gold standard for recording quality, this study shows that less expensive open-source acoustic recorders sampling at low frequency are viable for PAM where the cost of commercial devices is prohibitive. However, it should be noted that longer survey periods or use of multiple units are often required to obtain comparable data.

This paper presents a socially-engaged arts approach to exploring the variety and specificity of cultural benefits urban park-users associated with its treescape during the COVID-19 pandemic. Drawing on the cultural ecosystem services framework, cultural benefits are understood in terms of positive impacts to ‘experiences’ (our relational interactions with the environment), ‘capabilities’ (our knowledge and abilities in relation to environmental interaction) and ‘identities’ (our perceptions of our relationship with the environment). The research captured evidence of a broad range of people’s interactions with the human and non-human world, whilst opening up an inclusive space for respondents to reflect on and share feelings about the significance of these experiences. The methods employed attracted a range of ‘quieter voices’ to participate, particularly more vulnerable park-users. They also attended to the multiple levels at which people connected with treescapes during this time, from less conscious material engagements to more emotionally and culturally driven transactions. This case study research highlights the important role of the park’s treescape in supporting people to feel better during the COVID-19 crisis and their cultural associations and ties to it. However, it also explores feelings of concern for and perceived lack of influence over this valued resource as potentially disbenefiting wellbeing. It identifies experiences of environmental anxiety, emerging from a lack of certainty over and agency within urban green spaces and treescapes and the benefits they can provide. It concludes that management of treescapes and greenspaces should be sensitive to impacts on environmental emotion.

Temperate woodland vegetation is initially deter-mined by spatiotemporal and historical factors,mediated by complex biotic interactions.However,catastrophic events such as disease outbreaks(e.g.,sweet chestnut blight,ash dieback),infestations of insect pests,and human-accelerated climate change can create canopy gaps due to systematic decline in,or loss of,tree species that was once an impor-tant part of the canopy.Resultant cascade effects have the potential to alter the composition of woodland ecosystems quickly and radically,but inherent lag times make primary research into these effects challenging.Here,we explore change in woodland vegetation at 10 sites in response to can-opy opening using the Elm Decline,a rapid loss of Ulmus in woodlands across northwestern Europe～5800 years ago that coexisted alongside other stressors such as increasing human activity,as a palaeoecological analogue.For arboreal taxa,community evenness significantly decreased,within-site turnover significantly increased,and richness remained unchanged.Changes in arboreal taxa were highly site-spe-cific but there was a substantial decline in woody climbing taxa,especially Hedera(ivy),across the majority of sites.For shrub taxa,richness significantly increased but evenness and turnover remained consistent.Interestingly,however,there was a significant increase in abundance of shrubs at 70％of sites,including Calluna(heather),Ilex(holly)and Corylus(hazel),suggesting structural change.Surprisingly,there was no change in richness,evenness or turnover for herb taxa,possibly because change was highly variable spa-tially.However,there was a marked uptick in the disturbance indicator Plantago(plantain).Overall,these findings suggest that woodlands with sustained reduction in,or loss of,a tree species that once formed an important part of the canopy has profound,but often spatially idiosyncratic,impacts on veg-etation diversity(richness),composition(evenness),stabil-ity(turnover),and on abundance of specific taxa,especially within the shrub layer.Use of this palaeoecological ana-logue,which was itself complicated by cooccurring changes in human activity,provides a valuable empirical insight into possible cascade effects of similar change in canopy opening in contemporary settings,including Ash Dieback.

Effective monitoring, management and conservation of wildlife axiomatically depend on accurate data but causes of variation, including inter‐observer variation, are rarely explicitly quantified. Here, under controlled conditions, we demonstrate considerable variation in detection, identification and enumeration of (theoretically) readily identifiable African mammals at a reserve with a known species assemblage. Detection: 97.8% of sightings were missed by ≥1 observer; frequency of detection was affected by observer ID, detection distance, visibility and animal group size. Identification: just 3/14 species were identified consistently at all sightings. Enumeration: lack of consensus for 60.5% of sightings; consensus likelihood was affected by visibility and group size.

Effective wildlife conservation relies on surveying species, yet traditional methods are often financially and labour-intensive. Leveraging social media posts is potentially a more efficient alternative. Here, community assemblages derived from mammal sightings in Kruger National Park, South Africa, over the same four-month period, were compared for five social media platforms: Instagram, X (formerly Twitter), Facebook, Flickr, and Kruger Latest Sightings (a bespoke platform). In total, 7860 unique sightings of 45 species were recorded. Latest Sightings was the most comprehensive and consistent platform (37 species, 82

Riparian ecosystems are valuable habitats for bats, due to high densities of emergent aquatic invertebrates that provide high-quality feeding areas. Throughout Europe, decline and extirpation of European beaver (Castor fiber) has been a key driver in the decline of the extent and quality of riparian habitat, and thus recolonisation has considerable potential to restore degraded areas. Previous research has shown that beaver-modified ecosystems can support more bats, but the assumed causal link (an increase in invertebrate prey) has not been tested. Here, we study bat activity and richness/abundance of aquatic macroinvertebrates at a site where free-roaming beavers have colonised naturally from a nearby (unofficial) release location versus a nearby control site with very similar hydrology and habitat matrix. Bat activity was recorded using walked activity transects and fixed-point acoustic detectors. Although bat species richness was similar, bat activity was substantially and significantly higher at the beaver site versus the control site overall (42.7%) and for four specific taxa: soprano pipistrelle (Pipistrellus pygmaeus; 8.4%), Daubenton's (Myotis daubentonii; 46.5%), common pipistrelle (Pipistrellus pipistrellus; 110.4%) and common noctule (Nyctalus noctula; 170.9%). Richness of the larvae/nymphs of emergent aquatic macroinvertebrate species known to form an important part of bat diet was 205.5% higher at the beaver site compared to the control site, while abundance was 817.4% higher. Overall bat activity was also linked to habitat (lotic > lentic), survey method (transect > fixed-point), and season (May-July peak). This is the first evidence of a likely causal link between beaver recolonisation and bat activity via an increase in abundance of key aquatic invertebrate prey and we recommended that benefits of beaver reintroduction on bats be considered in future feasibility studies.

Nestboxes are commonly used to increase the number and quality of nest sites available to birds that usually use tree cavities and are considered an important conservation intervention. Although usually safer than natural cavities, birds nesting in simple, unmodified wooden nestboxes remain at risk of depredation. Accordingly, numerous design and placement modifications have been developed to ‘predator-proof’ nestboxes. These include: (1) adding metal plates around entrance holes to prevent enlargement; (2) affixing wire mesh to side panels; (3) deepening boxes to increase distance to nest cup; (4) creating external entrance ‘tunnels’ or internal wooden ledges; (5) using more robust construction materials; (6) developing photosensitive shutters to exclude nocturnal predators; (7) using baffles to block climbing mammals; and (8) regular replacement and relocation. However, the benefits and costs of these modifications are not always well understood. In this global review, we collate information on predator-proofing avian nestboxes designed for tree cavity-nesting birds to assess the efficacy of techniques for different predators (mammalian, avian, and reptilian) in different contexts. We critique the potential for modifications to have unintended consequences—including increasing nest building effort, altering microclimate, reducing provisioning rate, and elevating ectoparasite and microbial loads—to identify hidden costs. We conclude by highlighting remaining gaps in knowledge and providing guidance on optimal modifications in different contexts.

Widely available and inexpensive mobile phone applications offer users, whether professional ecologists or interested amateurs, the potential for simple and rapid automated identification of species, without the need to use field guides and identification keys. The increasing accuracy of machine learning is well established, but it is currently unclear if, and under what circumstances, free-to-use mobile phone applications are accurate for identifying plants to species level in real-world field conditions. We test five popular and free identification applications for plants using 857 professionally identified images of 277 species from 204 genera. Across all applications, 85% of images were identified correctly in the top five suggestions, and 69% were correct with the first suggestion. Plant type (woody, forbs, grasses, rushes/sedges, ferns/horsetails) was a significant determinant of identification performance for each application. For some applications, image saliency was also important; exposure and focus were not significant. Applications performed well, with at least one of the three best-performing applications identifying 96% of images correctly as their first suggestion. We conclude that, subject to some caveats, free phone-based plant identification applications are valid and useful tools for those wanting rapid identification and for anyone wanting to engage with the natural world. Read the free Plain Language Summary for this article on the Journal blog.

1. Socio-ecological research collaborations between artists, natural and social scientists, and with the humanities more broadly, have increased significantly in recent years. This has been aided by increased investment by funding bodies such as UK Research and Innovation and others internationally in projects designed to encourage cross-disciplinary partnerships.2. Within socio-ecological research and beyond it, despite some success stories, there is still a lack of awareness in 'the sciences' regarding how 'the arts' undertake their own forms of enquiry into the world. Further, different terminology and language used by different disciplines can cause confusion and misunderstanding, potentially leading to a reluctance to work collaboratively.3. In this paper, we discuss diversity within the arts as a discipline and seek to clarify various terminologies being used in both the arts and sciences to characterise joint working in research projects.4. Drawing on a series of semi-structured interviews and a workshop with artists and natural and social scientists with experience of collaboration in socio-ecological research, we compare understandings and expectations and reflect on the implications for funding bodies, institutions, artists and scientists which are widely applicable across different research contexts.

ABSTRACT Capsule Abundance of haematophagous ectoparasites in woodland passerine nests is influenced by complex interactions between nest box design, bird species, amount of nesting material and nest composition. Aims To analyse ectoparasite abundance relative to nest box design (old wooden nest boxes present for over 2 years versus new wooden nest boxes of the same dimensions vs deep wooden nest boxes designed to reduce predation risk) and bird species (Blue Tit Cyanistes caeruleus, Great Tit Parus major, Pied Flycatcher Ficedula hypoleuca and Eurasian Nuthatch Sitta europaea). The potential influence of amount of nesting material and nest composition was also studied. Methods After fledging, nests were collected from nest boxes. Ectoparasites and nest materials were identified and quantified. Generalized linear modelling was used to examine the influence of nest box design, bird species, amount of nest material and nest composition on ectoparasite loads. Akaike’s Information Criterion was used to select optimal models. Results Abundance of Hen Fleas Ceratophyllus gallinae and parasitic Blowfly Protocalliphora was significantly higher for deep nest boxes than nest boxes of standard dimensions. Old nest boxes had significantly higher loads than new nest boxes, despite thorough cleaning between breeding seasons. Hen Flea abundance was highest in Eurasian Nuthatch nests. Blowfly abundance was highest in Pied Flycatcher nests. Abundances of both fleas and blowfly were positively related to nest mass and amount of animal hair in the nest, and, for parasitic Blowfly, were negatively related to the amount of tree bark. Conclusion Ectoparasite load depends not only on bird species but also nest box design and nesting material. We recommend: (1) nest boxes are regularly replaced to reduce parasite load; (2) deep nest boxes are not used as the large nests constructed not only remove anti-predator benefits of eggs/chicks being harder to reach but are also associated with high haematophagous ectoparasite loads.

Engaging people with controversial and emotive issues that are inherently complex is challenging, especially when those issues are multifaceted or multidisciplinary, span environmental, economic, social, and political contexts, are global in scope, or where circumstances and implications differ spatially. Teaching such issues requires teachers and learners to navigate a challenging landscape of nuance and conflicting perspectives; immersive place-based learning might facilitate more meaningful exploration of such issues, but this has not previously been studied. In a multi-institution international study, we surveyed 164 participants (12 groups; 9 institutions) before and after an immersive fieldtrip in South Africa to assess perceptions on contentious issues. Wilcoxon signed-rank tests showed that participants had statistically significant opinion shifts for 17 of 18 statements, including those where change was likely due to improved knowledge or indicative of deeper attitude shifts. Generalised Linear Modelling revealed that propensity for opinion shifts was not influenced by demographics (age, gender), location (country of origin) or trip type (formal or informal). We conclude that in an increasingly complex world, context-relevant immersive experiences that facilitate deep learning by providing opportunities to explore contentious issues in situ are an ever-more valuable tool, not just for attainment but also for developing personal perspectives and as a catalyst for societal transitions.

Given the increasing obligation to elevate animal welfare beyond minimum expectations, zoos need robust mechanisms to monitor physical activity and species-appropriate behaviours. This is not without challenge as animal behaviour can vary seasonally or be influenced by external factors such as weather and visitors. In theory, keepers who work with animals year-round are ideally situated to collect behavioural data. However, time pressure often means any data collection is minimal, ad-hoc, or skewed towards particular times of day. Here we compare infrequent low-intensity ad-hoc data on activity and behaviour of zoo-housed Humboldt penguins (Spheniscus humboldti) collected by keepers with simultaneous high-intensity systematic data collected by a researcher. Once out-of-sight records (more frequent in researcher data) were excluded from both datasets, we found excellent agreement between keeper and researcher data. After this validation step, we collected extensive multi-year keeper data (diurnal hourly scan samples; 262 days across a 5-year study period) to assess potential influence of time of day, time of year, visitors, and weather on penguin behaviour. There were pronounced seasonal patterns: aquatic active behaviours were highest in autumn (38.9% in October) compared to early spring and late summer (February = 16.7%; August = 19.4%). Time of day was also important: terrestrial inactive behaviours increased throughout the day and this pattern was significantly more pronounced in summer than winter. Aquatic active behaviours were more frequent during warmer/sunnier conditions compared to cooler/wetter conditions, while the reverse was true for terrestrial inactive behaviours. Birds were much less active on days when the site was closed to visitors. Overall, we demonstrate that where keeper-collected data are validated and found to be robust and representative, there are real opportunities to study long-term behavioural patterns that can quickly translate to tangible husbandry benefits and improved animal welfare.

Although white rhinoceros (Ceratotherium simum) are common in captivity, few behavioral studies have been conducted and there is seemingly no research for immersive exhibits where potential for visitor effects is high. Moreover, little information exists on possible effects of weather and temperature on rhino outside their native range. Here we analyze 14,501 observations of rhino in a drive-through enclosure. Data were collected by researchers (n = 12,160 datapoints) and keepers (n = 2341 datapoints) over a 4-month period. We aimed to: (1) quantify behavior using detailed researcher-collected data and contemporaneous but ad hoc keeper-collected data; (2) compare datasets statistically; (3) establish effects of visitors, temperature, and weather on behavior; and (4) assess the influence of visitors on similarity of researcher/keeper datasets. Activity budgets were similar to the wild and the single previous study from a traditional (nondrive-through) enclosure. There was some discrepancy in activity budgets between researcher and keeper data due to significant differences in recorded frequency of two rare behaviors (horn rub; social interaction) and two behaviors that could be easily confused (grazing vs. standing with head-down): recording of other behaviors matched well. Weather and temperature affected behavior, with rhino becoming more sedentary (-locomotion, grazing; +resting, standing, and sedentary eating of hay) on hot/sunny days compared to cool/wet days. The number of visitor vehicles had a fairly negligible effect but resting was lower on busy days, possibly as vigilance increased. The match between researcher/keeper datasets was lowest when visitor numbers were high, suggesting visitors might affect keeper ability to accurately record behavior.

The phenology of many species shows strong sensitivity to climate change; however, with few large scale intra-specific studies it is unclear how such sensitivity varies over a species' range. We document large intra-specific variation in phenological sensitivity to temperature using laying date information from 67 populations of two co-familial European songbirds, the great tit (Parus major) and blue tit (Cyanistes caeruleus), covering a large part of their breeding range. Populations inhabiting deciduous habitats showed stronger phenological sensitivity than those in evergreen and mixed habitats. However, populations with higher sensitivity tended to have experienced less rapid change in climate over the past decades, such that populations with high phenological sensitivity will not necessarily exhibit the strongest phenological advancement. Our results show that to effectively assess the impact of climate change on phenology across a species' range it will be necessary to account for intra-specific variation in phenological sensitivity, climate change exposure, and the ecological characteristics of a population.

Identifying the environmental drivers of variation in fitness-related traits is a central objective in ecology and evolutionary biology. Temporal fluctuations of these environmental drivers are often synchronized at large spatial scales. Yet, whether synchronous environmental conditions can generate spatial synchrony in fitness-related trait values (i.e., correlated temporal trait fluctuations across populations) is poorly understood. Using data from long-term monitored populations of blue tits (Cyanistes caeruleus, n = 31), great tits (Parus major, n = 35), and pied flycatchers (Ficedula hypoleuca, n = 20) across Europe, we assessed the influence of two local climatic variables (mean temperature and mean precipitation in February-May) on spatial synchrony in three fitness-related traits: laying date, clutch size, and fledgling number. We found a high degree of spatial synchrony in laying date but a lower degree in clutch size and fledgling number for each species. Temperature strongly influenced spatial synchrony in laying date for resident blue tits and great tits but not for migratory pied flycatchers. This is a relevant finding in the context of environmental impacts on populations because spatial synchrony in fitness-related trait values among populations may influence fluctuations in vital rates or population abundances. If environmentally induced spatial synchrony in fitness-related traits increases the spatial synchrony in vital rates or population abundances, this will ultimately increase the risk of extinction for populations and species. Assessing how environmental conditions influence spatiotemporal variation in trait values improves our mechanistic understanding of environmental impacts on populations.

In southern Africa, climate change is increasing drought severity and duration. One way to mitigate drought-induced food shortages is supplementary feeding, but ungulate responses are largely unstudied. Here, we analyse similar to 250 h of video footage from camera traps at feeding stations in South Africa during a year when rainfall was similar to 50% of the decadal average. Feeding station usage was highly uneven: eland (Taurotragus oryx), greater kudu (Tragelaphus strepsiceros), sable (Hippotragus niger) and white rhino (Ceratotherium simum) were over-represented relative to their local population; and impala (Aepyceros melampus), blesbok (Damascus dorxas phillipsi) and zebra (Equus quagga) were under-represented. In total, 81% of individuals at feeding stations fed (0%-90% for different species). Guarding behaviour was observed, especially by eland and kudu, potentially excluding other (smaller) species. Overall, 17% visits involving >= 2 individuals involved intraspecific aggression (highest for sable (40%) and eland (27%); low/absent otherwise); there were very few (n = 6) inter-specific encounters. Lucerne and teff hay were consumed more frequently than Boskos (pellets), which was favoured by non-target primates and ungulates showing guarding behaviour. As benefits of feeding are unequal, and can encourage guarding and intraspecific aggression, we suggest feeding should be done with caution, especially if the aim is to benefit specific species.